figure_1
Vincent Fischer
16 11 2022
suppressPackageStartupMessages({
library(epiwraps)
library(rtracklayer)
library(GenomicRanges)
library(ggplot2)
library(ComplexHeatmap)
library(chromVAR)
library(motifmatchr)
library(memes)
library(MotifDb)
library(universalmotif)
library(TFBSTools)
library(AnnotationHub)
library(RColorBrewer)
library(viridis)
library(viridisLite)
library(ggrepel)
library(magick)
library(ggpubr)
library(cowplot)
})## See system.file("LICENSE", package="MotifDb") for use restrictions.theme_set(theme_bw())
set.seed(123)1 A
1.1 load and prepare Data
load signal files load Veh track for NF,Mono and Full; and give names
bw_ATAC_mono <- list.files(path = "../Dex_Veh/tracks/", pattern ="*Veh.(.*)mono.bw",full.names = TRUE )
bw_ATAC_nf <- list.files(path = "../Dex_Veh/tracks/", pattern ="*Veh.(.*)NF.bw",full.names = TRUE )
bw_ATAC_full <- list.files(path = "../Dex_Veh/tracks/", pattern ="*Veh.(.*)full.bw",full.names = TRUE )
names(bw_ATAC_mono) <- c("ATAC_Mono_1","ATAC_Mono_2")
names(bw_ATAC_nf) <- c("ATAC_NF_1","ATAC_NF_2")
names(bw_ATAC_full) <- c("ATAC_Full_1","ATAC_Full_2")import 5fC sites and getting unique 5fC and no5fC sites
sites_5fC <- readRDS("../object_resources/5fC/sites_5fC.rds")
sites_no5fC <- readRDS("../object_resources/5fC/sites_no5fC.rds")load TSS with and without 5fC
TSS_2KB_5fC <- readRDS("../object_resources/5fC/TSS_2KB_5fC.rds")
TSS_2KB_no5fC <- readRDS("../object_resources/5fC/TSS_2KB_no5fC.rds")generation of normalization factors
bg_nf <- bwNormFactors(c(bw_ATAC_nf,bw_ATAC_mono,bw_ATAC_full), method = "background")## Warning: bwNormFactors is deprecated, please gradually switch to
## `getNormFactors`.## Comparing coverage in random regions...sample no5fC regions to match
sampled_no5fC <- TSS_2KB_no5fC[sample(c(1:length(TSS_2KB_no5fC)),size = length(TSS_2KB_5fC))]
list_5fC_no5fC <- unlist(GRangesList(TSS_2KB_5fC,sampled_no5fC))cluster regions 5fC top and no5fC bottom
# cluster regions 5fC top and no5fC bottom
i <- integer(length(list_5fC_no5fC))
i[1:(length(list_5fC_no5fC)/2)] <- "TSS with 5fC"
i[((length(list_5fC_no5fC)/2)+1):length(list_5fC_no5fC)] <- "TSS with no 5fC"
saveRDS(i,"../object_resources/figure_1_meta/cluster_5fC_no5fC.LIST.rds")mycolors <- c("TSS with 5fC"=brewer.pal(n=9,"Set1")[1], "TSS with no 5fC"=brewer.pal(n=9,"Set1")[2])
m_5fC_no5fC <- signal2Matrix(c(bw_ATAC_nf,bw_ATAC_mono,bw_ATAC_full),w = 2L, regions = list_5fC_no5fC,type = "center")## Reading ../Dex_Veh/tracks//Veh4.NF.bw## Reading ../Dex_Veh/tracks//Veh5.NF.bw## Reading ../Dex_Veh/tracks//Veh4.mono.bw## Reading ../Dex_Veh/tracks//Veh5.mono.bw## Reading ../Dex_Veh/tracks//Veh4.full.bw## Reading ../Dex_Veh/tracks//Veh5.full.bwm_5fC_no5fC_bg <- rescaleSignalMatrices(m_5fC_no5fC,bg_nf)
#merging replicates in the signal matrices
sm2 <- list( Nuc.Free =mergeSignalMatrices(m_5fC_no5fC_bg[1:2],aggregation = "mean"), Mononucleosome=mergeSignalMatrices(m_5fC_no5fC_bg[3:4],aggregation = "mean"), Full=mergeSignalMatrices(m_5fC_no5fC_bg[5:6],aggregation = "mean") )saveRDS(sm2,"../object_resources/figure_1_meta/matrix_mm.rds")
m_mm <- readRDS("../object_resources/figure_1_meta/matrix_mm.rds")1.2 generate A
mycolors <- c("TSS with 5fC"=brewer.pal(n=9,"Set1")[1], "TSS with no 5fC"=brewer.pal(n=9,"Set1")[2])
i <- readRDS("../object_resources/figure_1_meta/cluster_5fC_no5fC.LIST.rds")
p1 <- epiwraps::plotEnrichedHeatmaps((m_mm), row_split = i,scale_title = "Backg.\nnorm.\ndensity", axis_name = c("-2kb","TSS","+2kb"), colors = rocket(100), mean_color = mycolors, mean_scale_side = "left", top_annotation = FALSE,use_raster = TRUE)
p1
saveRDS(p1,"../object_resources/figure_1_meta/p1.rds")2 B
2.1 prepare Data
generate individual matrices for 5fC and no5fC
# with 5fC
m_5fC <- signal2Matrix(c(bw_ATAC_nf,bw_ATAC_mono,bw_ATAC_full),w = 10L, regions = TSS_2KB_5fC)## Reading ../Dex_Veh/tracks//Veh4.NF.bw## Reading ../Dex_Veh/tracks//Veh5.NF.bw## Reading ../Dex_Veh/tracks//Veh4.mono.bw## Reading ../Dex_Veh/tracks//Veh5.mono.bw## Reading ../Dex_Veh/tracks//Veh4.full.bw## Reading ../Dex_Veh/tracks//Veh5.full.bwm_5fC_bg <- rescaleSignalMatrices(m_5fC,bg_nf)
# without 5fC
m_no5fC <- signal2Matrix(c(bw_ATAC_nf,bw_ATAC_mono,bw_ATAC_full),w = 10L, regions = TSS_2KB_no5fC)## Reading ../Dex_Veh/tracks//Veh4.NF.bw## Reading ../Dex_Veh/tracks//Veh5.NF.bw## Reading ../Dex_Veh/tracks//Veh4.mono.bw## Reading ../Dex_Veh/tracks//Veh5.mono.bw## Reading ../Dex_Veh/tracks//Veh4.full.bw## Reading ../Dex_Veh/tracks//Veh5.full.bwm_no5fC_bg <- rescaleSignalMatrices(m_no5fC,bg_nf)aggregate signal
# with the matrices we generate dataframes with mean signal per position
d_5fC <- meltSignals(m_5fC_bg,trim = c(0.02,0.98))
d_no5fC <- meltSignals(m_no5fC_bg,trim = c(0.02,0.98))
# we merge both dataframes
dmerge <- dplyr::bind_rows(list("5fC"=d_5fC, "no5fC"=d_no5fC), .id="regions")
dmergesaveRDS(dmerge,"../object_resources/figure_1_meta/dmerge_1.rds")# we load the merged dataframe
dmerge <- readRDS("dmerge_1.rds")2.2 generate figure B
# we plot the aggregate signals by region and fragment size distribution
p2 <- plot_grid(
ggplot(dmerge[dmerge$sample == "ATAC_NF_1"|dmerge$sample == "ATAC_NF_2",], aes(position, mean, colour=regions, group = regions, fill = regions)) +
geom_vline(xintercept=0, linetype="dashed") +
labs(x="TSS", y="mean background norm. density")+
stat_summary(geom="ribbon", alpha=0.3, colour = NA ) + stat_summary(fun=mean, geom="line", size = 0.6)+
ggtitle("Nuc. Free")+ scale_color_brewer(palette="Set1")+scale_fill_brewer(palette="Set1")+ylim(0,0.1)+theme(text = element_text(size = 10),legend.position = "none"),
ggplot(dmerge[dmerge$sample == "ATAC_Mono_1"|dmerge$sample == "ATAC_Mono_2",], aes(position, mean, colour=regions, group = regions, fill = regions)) +
geom_vline(xintercept=0, linetype="dashed") +
stat_summary(geom="ribbon", alpha=0.3, colour = NA ) + stat_summary(fun=mean, geom="line", size = 0.6)+
labs(x="TSS")+
ggtitle("Mononucleosome")+ scale_color_brewer(palette="Set1")+scale_fill_brewer(palette="Set1")+ylim(0,0.1)+theme(text = element_text(size = 10),axis.title.y = element_blank(),legend.position = "none",axis.text.y = element_blank()),
ggplot(dmerge[dmerge$sample == "ATAC_Full_1"|dmerge$sample == "ATAC_Full_2",], aes(position, mean, colour=regions, group = regions, fill = regions)) +
geom_vline(xintercept=0, linetype="dashed") +
stat_summary(geom="ribbon", alpha=0.3, colour = NA ) + stat_summary(fun=mean, geom="line", size = 0.6)+
xlab("TSS")+
ggtitle("Full")+ scale_color_brewer(palette="Set1")+scale_fill_brewer(palette="Set1")+ylim(0,0.1)+theme(text = element_text(size = 10),axis.title.y = element_blank(),axis.text.y = element_blank()),
ncol = 3, nrow = 1,rel_widths = c(1.15,1,1.2)
) ## Warning: Using `size` aesthetic for lines was deprecated in ggplot2 3.4.0.
## ℹ Please use `linewidth` instead.
## This warning is displayed once every 8 hours.
## Call `lifecycle::last_lifecycle_warnings()` to see where this warning was
## generated.## No summary function supplied, defaulting to `mean_se()`
## No summary function supplied, defaulting to `mean_se()`
## No summary function supplied, defaulting to `mean_se()`p2
saveRDS(p2,"../object_resources/figure_1_meta/p2.rds")3 C
3.1 load and prepare data
previous data was generated with methylation data from an unpublished source therefore it is now repeated with published work (rando)
import unique 5fC sites in proximity to TSS
unique_5fC_sites_in_TSS <- readRDS("../object_resources/5fC/unique_5fC_sites_in_TSS.rds")import genome fasta file (GRCm38)
genome <- Rsamtools::FaFile("../object_resources/Mus_musculus.GRCm38.dna_sm.primary_assembly.fa")
genome## class: FaFile
## path: ../object_resources/Mus_musculus.GRCm38.dna_sm.primary_assembly.fa
## index: ../object_resources.../Mus_musculus.GRCm38.dna_sm.primary_assembly.fa.fai
## gzindex: ../object_resourc.../Mus_musculus.GRCm38.dna_sm.primary_assembly.fa.gzi
## isOpen: FALSE
## yieldSize: NAloading Rando data methylation levels at promoters and perform liftover (PL)
methylation_mm9 <- readRDS("../object_resources/GSE75323_sperm_prom_methylation.mm9.rds")
methylation_mm9 <- GRanges(methylation_mm9)
seqlevelsStyle(methylation_mm9)## [1] "UCSC"mm9_mm10_chain <- import.chain("../object_resources//mm9ToMm10.over.chain")
methylation_total <- liftOver(methylation_mm9,chain = mm9_mm10_chain)
methylation_total <- unlist(methylation_total)
seqlevelsStyle(methylation_total) <- "Ensembl"
saveRDS(methylation_total,"../object_resources/methylation_total_grcm38.GR.rds")load GRCm38 sperm methylation data and check for promoters with methylation score >0.5
#methylation_total <- readRDS("../object_resources/methylation_rando_grcm38.GR.rds")
methyl_promoters <- methylation_total[methylation_total$meanMeth>0.5]load motifs and check select non-redundant TFs
core <- read_meme(file = "../object_resources/HOCOMOCOv11_core_MOUSE_mono_meme_format.meme") # needs to be the right path
#full <- read_meme(file = "HOCOMOCOv11_full_MOUSE_mono_meme_format.meme")
# original function PL
summary_core <- universalmotif::summarise_motifs(core)
core <- setNames(object = core,nm = summary_core$name)
motifs <- core
modf <- data.frame(row.names=summary_core$name,
TF=gsub("_MOUSE.H11MO...+","",summary_core$name),
grade=gsub(".+\\.","",summary_core$name),
motif_rank = gsub("[^01]+","",gsub(".*MO.",replacement = "",summary_core$name))) #gsub("[^ro]+", "", str1)
modf <- modf[order(modf$TF,modf$grade,modf$motif_rank),]
modf <- modf[!duplicated(modf$TF),]
modf <- modf[!modf$grade == "D",]
motifs <- setNames(universalmotif::convert_motifs(motifs[row.names(modf)]), modf$TF)peak_centers <- resize(unique_5fC_sites_in_TSS,fix = "center",width = 200)
peak_seqs <- memes::get_sequence(dropSeqlevels(peak_centers,"MT",pruning.mode = "coarse"), genome)# dropping MT due to its absence in reference genome file
peak_centers <- reduce(promoters(methyl_promoters, upstream = 1000, downstream = 100))
peak_seqs_methyl_prom <- memes::get_sequence(dropSeqlevels(peak_centers,"MT",pruning.mode = "coarse"), genome)# dropping MT due to its absence in reference genome file
ame <- memes::runAme(peak_seqs,control = peak_seqs_methyl_prom , database=convert_motifs(motifs,class = "universalmotif-universalmotif"), meme_path="/common/meme/bin/")
head(ame)ame$motif <- basename(ame$motif_db)
saveRDS(ame,"../object_resources/figure_1_meta/motif_enrichment_5fC_5mC_promoter.rds")load sample
#ame<- readRDS("../object_resources/figure_1_meta/motif_enrichment_rando.rds")
ame <- ame[!ame$fp==0,] #TFs with no false positive are removed since no fold change can be calculated3.2 generate figure C
# results with no false positive were removed
p3 <- ggplot(ame, aes(log2(((tp/pos))/((fp/neg))), -log10(adj.pvalue))) +
geom_point(alpha=0.7, size =3) + geom_text_repel(data=head((ame),9), aes(label=motif),max.overlaps = 20, size = 5) +
labs(x="log2FE")+ xlim(c(0,6.5))+theme(text = element_text(size = 12))#+ ggtitle("5fC sites in TSS proximity compared to methylated promoter regions")
p3
4 combine all figures
#plot_grid(grid.grabExpr(ComplexHeatmap::draw(p1)), nrow=1, rel_widths=c(5,3))
pp <- plot_grid(plot_grid(grid.grabExpr(ComplexHeatmap::draw(p1)),p3,labels = c("A","C"),rel_widths = c(3,2)),p2,labels = c(NA,"B"),nrow = 2,rel_heights = c(3,2))
pp## Warning: Removed 1 rows containing missing values (`geom_text()`).
# now add the title
title <- ggdraw() +
draw_label(
"Figure 1",
fontface = 'bold',
x = 0,
hjust = 0
) +
theme(
# add margin on the left of the drawing canvas,
# so title is aligned with left edge of first plot
plot.margin = margin(0, 0, 0, 7)
)
pp2 <- plot_grid(
title, pp,
ncol = 1,
# rel_heights values control vertical title margins
rel_heights = c(0.1, 2)
)## Warning: Removed 1 rows containing missing values (`geom_text()`).pp2
pdf("figure_1_meta.pdf", width=10, height=7.5)
pp2
dev.off()## png
## 2sessionInfo()## R version 4.0.3 (2020-10-10)
## Platform: x86_64-pc-linux-gnu (64-bit)
## Running under: Ubuntu 18.04 LTS
##
## Matrix products: default
## BLAS: /usr/lib/x86_64-linux-gnu/openblas/libblas.so.3
## LAPACK: /usr/lib/x86_64-linux-gnu/libopenblasp-r0.2.20.so
##
## locale:
## [1] LC_CTYPE=C.UTF-8 LC_NUMERIC=C LC_TIME=C.UTF-8
## [4] LC_COLLATE=C.UTF-8 LC_MONETARY=C.UTF-8 LC_MESSAGES=C.UTF-8
## [7] LC_PAPER=C.UTF-8 LC_NAME=C LC_ADDRESS=C
## [10] LC_TELEPHONE=C LC_MEASUREMENT=C.UTF-8 LC_IDENTIFICATION=C
##
## attached base packages:
## [1] grid parallel stats4 stats graphics grDevices utils
## [8] datasets methods base
##
## other attached packages:
## [1] cowplot_1.1.1 ggpubr_0.4.0 magick_2.7.3
## [4] ggrepel_0.9.3 viridis_0.6.2 viridisLite_0.4.1
## [7] RColorBrewer_1.1-3 AnnotationHub_2.22.1 BiocFileCache_1.14.0
## [10] dbplyr_2.1.1 TFBSTools_1.28.0 universalmotif_1.8.5
## [13] MotifDb_1.32.0 Biostrings_2.58.0 XVector_0.30.0
## [16] memes_0.99.11 motifmatchr_1.12.0 chromVAR_1.12.0
## [19] ggplot2_3.4.2 rtracklayer_1.50.0 epiwraps_0.99.62
## [22] EnrichedHeatmap_1.29.2 ComplexHeatmap_2.15.1 GenomicRanges_1.42.0
## [25] GenomeInfoDb_1.26.7 IRanges_2.24.1 S4Vectors_0.28.1
## [28] BiocGenerics_0.36.1
##
## loaded via a namespace (and not attached):
## [1] rappdirs_0.3.3 R.methodsS3_1.8.1
## [3] tidyr_1.2.0 bit64_4.0.5
## [5] knitr_1.42 DelayedArray_0.16.3
## [7] R.utils_2.11.0 data.table_1.14.8
## [9] rpart_4.1.16 KEGGREST_1.30.1
## [11] RCurl_1.98-1.6 AnnotationFilter_1.14.0
## [13] doParallel_1.0.17 generics_0.1.3
## [15] GenomicFeatures_1.42.3 RSQLite_2.2.11
## [17] bit_4.0.5 tzdb_0.3.0
## [19] xml2_1.3.3 httpuv_1.6.9
## [21] SummarizedExperiment_1.20.0 assertthat_0.2.1
## [23] DirichletMultinomial_1.32.0 xfun_0.38
## [25] hms_1.1.1 jquerylib_0.1.4
## [27] evaluate_0.20 promises_1.2.0.1
## [29] fansi_1.0.4 progress_1.2.2
## [31] caTools_1.18.2 DBI_1.1.3
## [33] htmlwidgets_1.5.4 purrr_1.0.1
## [35] ellipsis_0.3.2 dplyr_1.1.1
## [37] backports_1.4.1 annotate_1.68.0
## [39] biomaRt_2.46.3 MatrixGenerics_1.2.1
## [41] vctrs_0.6.1 Biobase_2.50.0
## [43] ensembldb_2.14.1 Cairo_1.6-0
## [45] abind_1.4-5 cachem_1.0.7
## [47] withr_2.5.0 Gviz_1.34.1
## [49] BSgenome_1.58.0 vroom_1.5.7
## [51] checkmate_2.1.0 GenomicAlignments_1.26.0
## [53] prettyunits_1.1.1 cluster_2.1.4
## [55] lazyeval_0.2.2 seqLogo_1.56.0
## [57] crayon_1.5.2 edgeR_3.32.1
## [59] pkgconfig_2.0.3 labeling_0.4.2
## [61] pkgload_1.3.2 ProtGenerics_1.22.0
## [63] nnet_7.3-17 rlang_1.1.0
## [65] lifecycle_1.0.3 miniUI_0.1.1.1
## [67] dichromat_2.0-0.1 rprojroot_2.0.3
## [69] matrixStats_0.63.0 Matrix_1.5-3
## [71] ggseqlogo_0.1 carData_3.0-5
## [73] base64enc_0.1-3 processx_3.5.3
## [75] GlobalOptions_0.1.2 png_0.1-7
## [77] rjson_0.2.21 bitops_1.0-7
## [79] splitstackshape_1.4.8 cmdfun_1.0.2
## [81] R.oo_1.24.0 blob_1.2.2
## [83] shape_1.4.6 stringr_1.5.0
## [85] readr_2.1.2 jpeg_0.1-10
## [87] rstatix_0.7.0 ggsignif_0.6.4
## [89] CNEr_1.26.0 scales_1.2.1
## [91] memoise_2.0.1 magrittr_2.0.3
## [93] plyr_1.8.8 zlibbioc_1.36.0
## [95] compiler_4.0.3 clue_0.3-60
## [97] Rsamtools_2.6.0 cli_3.6.1
## [99] ps_1.6.0 htmlTable_2.4.0
## [101] Formula_1.2-5 MASS_7.3-56
## [103] tidyselect_1.2.0 stringi_1.7.12
## [105] highr_0.10 yaml_2.3.7
## [107] askpass_1.1 locfit_1.5-9.4
## [109] latticeExtra_0.6-29 sass_0.4.1
## [111] VariantAnnotation_1.36.0 tools_4.0.3
## [113] circlize_0.4.15 rstudioapi_0.13
## [115] TFMPvalue_0.0.8 foreach_1.5.2
## [117] foreign_0.8-84 gridExtra_2.3
## [119] farver_2.1.1 digest_0.6.31
## [121] BiocManager_1.30.20 shiny_1.7.1
## [123] pracma_2.3.8 Rcpp_1.0.10
## [125] car_3.0-12 broom_0.7.12
## [127] BiocVersion_3.12.0 later_1.3.0
## [129] httr_1.4.2 AnnotationDbi_1.52.0
## [131] biovizBase_1.38.0 Rdpack_2.3
## [133] colorspace_2.1-0 brio_1.1.3
## [135] XML_3.99-0.9 splines_4.0.3
## [137] plotly_4.10.0 xtable_1.8-4
## [139] jsonlite_1.8.4 poweRlaw_0.70.6
## [141] GenomicFiles_1.26.0 UpSetR_1.4.0
## [143] testthat_3.1.2 R6_2.5.1
## [145] Hmisc_4.6-0 pillar_1.9.0
## [147] htmltools_0.5.5 mime_0.12
## [149] glue_1.6.2 fastmap_1.1.1
## [151] DT_0.22 BiocParallel_1.24.1
## [153] interactiveDisplayBase_1.28.0 codetools_0.2-19
## [155] utf8_1.2.3 lattice_0.21-8
## [157] bslib_0.3.1 tibble_3.2.1
## [159] curl_5.0.0 gtools_3.9.4
## [161] GO.db_3.12.1 openssl_2.0.0
## [163] survival_3.3-1 limma_3.46.0
## [165] rmarkdown_2.13 desc_1.4.2
## [167] munsell_0.5.0 GetoptLong_1.0.5
## [169] GenomeInfoDbData_1.2.4 iterators_1.0.14
## [171] reshape2_1.4.4 gtable_0.3.3
## [173] rbibutils_2.2.8